A Novel Approach to Design 3D Models in Medical Education.
Ontology highlight
ABSTRACT: Supplementary information
The online version contains supplementary material available at 10.1007/s40670-021-01262-6.
SUBMITTER: Rosen D
PROVIDER: S-EPMC7983107 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundThree-dimensional (3D) printing technology enables the translation of 2-dimensional (2D) medical imaging into a physical replica of a patient's individual anatomy and may enhance the understanding of congenital heart defects (CHD). We aimed to evaluate the usefulness of a spectrum of 3D-printed models in teaching CHD to medical students.ResultsWe performed a prospective, randomized educational procedure to teach fifth year medical students four CHDs (atrial septal defect (ASD, n = 74), ventricular septal defect (VSD, n = 50), coarctation of aorta (CoA, n = 118) and tetralogy of Fallot (ToF, n = 105)). Students were randomized into printing groups or control groups. All students received the same 20 min lecture with projected digital 2D images. The printing groups also manipulated 3D printed models during the lecture. Both groups answered an objective survey (Multiple-choice questionnaire) twice, pre- and post-test, and completed a post-lecture subjective survey. Three hundred forty-seven students were included and both teaching groups for each CHD were comparable in age, sex and pre-test score. Overall, objective knowledge improved after the lecture and was higher in the printing group compared to the control group (16.3 ± 2.6 vs 14.8 ± 2.8 out of 20, p < 0.0001). Similar results were observed for each CHD (p = 0.0001 ASD group; p = 0.002 VSD group; p = 0.0005 CoA group; p = 0.003 ToF group). Students' opinion of their understanding of CHDs was higher in the printing group compared to the control group (respectively 4.2 ± 0.5 vs 3.8 ± 0.4 out of 5, p < 0.0001).ConclusionThe use of 3D printed models in CHD lectures improve both objective knowledge and learner satisfaction for medical students. The practice should be mainstreamed.
| S-EPMC8424617 | biostudies-literature
Project description:Transfer learning from natural image to medical image has established as one of the most practical paradigms in deep learning for medical image analysis. However, to fit this paradigm, 3D imaging tasks in the most prominent imaging modalities (e.g., CT and MRI) have to be reformulated and solved in 2D, losing rich 3D anatomical information and inevitably compromising the performance. To overcome this limitation, we have built a set of models, called Generic Autodidactic Models, nicknamed Models Genesis, because they are created ex nihilo (with no manual labeling), self-taught (learned by self-supervision), and generic (served as source models for generating application-specific target models). Our extensive experiments demonstrate that our Models Genesis significantly outperform learning from scratch in all five target 3D applications covering both segmentation and classification. More importantly, learning a model from scratch simply in 3D may not necessarily yield performance better than transfer learning from ImageNet in 2D, but our Models Genesis consistently top any 2D approaches including fine-tuning the models pre-trained from ImageNet as well as fine-tuning the 2D versions of our Models Genesis, confirming the importance of 3D anatomical information and significance of our Models Genesis for 3D medical imaging. This performance is attributed to our unified self-supervised learning framework, built on a simple yet powerful observation: the sophisticated yet recurrent anatomy in medical images can serve as strong supervision signals for deep models to learn common anatomical representation automatically via self-supervision. As open science, all pre-trained Models Genesis are available at https://github.com/MrGiovanni/ModelsGenesis.
| S-EPMC7405596 | biostudies-literature
Project description:Medical AI has transformed modern medicine and created a new environment for future doctors. However, medical education has failed to keep pace with these advances, and it is essential to provide systematic education on medical AI to current medical undergraduate and postgraduate students. To address this issue, our study utilized the Unified Theory of Acceptance and Use of Technology model to identify key factors that influence the acceptance and intention to use medical AI. We collected data from 1,243 undergraduate and postgraduate students from 13 universities and 33 hospitals, and 54.3% reported prior experience using medical AI. Our findings indicated that medical postgraduate students have a higher level of awareness in using medical AI than undergraduate students. The intention to use medical AI is positively associated with factors such as performance expectancy, habit, hedonic motivation, and trust. Therefore, future medical education should prioritize promoting students' performance in training, and courses should be designed to be both easy to learn and engaging, ensuring that students are equipped with the necessary skills to succeed in their future medical careers.
| S-EPMC10637014 | biostudies-literature
Project description:While nutritional interventions are first-line therapy for many chronic diseases, most medical trainees receive minimal nutrition education, leaving them unprepared to address nutritional issues with patients. An interactive, single-session, virtual nutrition curriculum was taught online to 80 physician assistant (PA) students. Topics included plant-based nutrition, dietary history-taking and counseling, and culinary medicine. Students were surveyed before, immediately after, and four weeks after the curriculum to assess changes to nutrition-related knowledge, attitudes, confidence, and personal dietary behaviors. Seventy-three PA students (91%) completed the pre-survey, 76 (95%) completed the post-survey, and 42 (52.5%) completed the delayed post-survey. Knowledge scores increased immediately post-intervention (48.9% to 78.9%; p < 0.001) and persisted four weeks later (78.9% to 75.8%; p = 0.54). Post-intervention, students felt more confident in dietary history-taking (55% vs. 95%; p = 0.001) and nutrition counseling (53% vs. 84%; p = 0.003) and agreed that dietary changes alone could reverse type 2 diabetes (74% vs. 97%; p = 0.027) and coronary artery disease (66% vs. 92%; p = 0.039). Curricula using virtual teaching kitchens may be a scalable approach to nutrition education for medical trainees.
| S-EPMC10574391 | biostudies-literature
Project description:Rational compound design remains a challenging problem for both computational methods and medicinal chemists. Computational generative methods have begun to show promising results for the design problem. However, they have not yet used the power of three-dimensional (3D) structural information. We have developed a novel graph-based deep generative model that combines state-of-the-art machine learning techniques with structural knowledge. Our method ("DeLinker") takes two fragments or partial structures and designs a molecule incorporating both. The generation process is protein-context-dependent, utilizing the relative distance and orientation between the partial structures. This 3D information is vital to successful compound design, and we demonstrate its impact on the generation process and the limitations of omitting such information. In a large-scale evaluation, DeLinker designed 60% more molecules with high 3D similarity to the original molecule than a database baseline. When considering the more relevant problem of longer linkers with at least five atoms, the outperformance increased to 200%. We demonstrate the effectiveness and applicability of this approach on a diverse range of design problems: fragment linking, scaffold hopping, and proteolysis targeting chimera (PROTAC) design. As far as we are aware, this is the first molecular generative model to incorporate 3D structural information directly in the design process. The code is available at https://github.com/oxpig/DeLinker.
| S-EPMC7189367 | biostudies-literature
Project description:Recent advances in computer vision have shown promising results in image generation. Diffusion probabilistic models have generated realistic images from textual input, as demonstrated by DALL-E 2, Imagen, and Stable Diffusion. However, their use in medicine, where imaging data typically comprises three-dimensional volumes, has not been systematically evaluated. Synthetic images may play a crucial role in privacy-preserving artificial intelligence and can also be used to augment small datasets. We show that diffusion probabilistic models can synthesize high-quality medical data for magnetic resonance imaging (MRI) and computed tomography (CT). For quantitative evaluation, two radiologists rated the quality of the synthesized images regarding "realistic image appearance", "anatomical correctness", and "consistency between slices". Furthermore, we demonstrate that synthetic images can be used in self-supervised pre-training and improve the performance of breast segmentation models when data is scarce (Dice scores, 0.91 [without synthetic data], 0.95 [with synthetic data]).
| S-EPMC10163245 | biostudies-literature
Project description:IntroductionOur objective was to evaluate the effectiveness of hands-on training at a bedside ultrasound (US) symposium ("Ultrafest") to improve both clinical knowledge and image acquisition skills of medical students. Primary outcome measure was improvement in multiple choice questions on pulmonary or Focused Assessment with Sonography in Trauma (FAST) US knowledge. Secondary outcome was improvement in image acquisition for either pulmonary or FAST.MethodsProspective cohort study of 48 volunteers at "Ultrafest," a free symposium where students received five contact training hours. Students were evaluated before and after training for proficiency in either pulmonary US or FAST. Proficiency was assessed by clinical knowledge through written multiple-choice exam, and clinical skills through accuracy of image acquisition. We used paired sample t-tests with students as their own controls.ResultsPulmonary knowledge scores increased by a mean of 10.1 points (95% CI [8.9-11.3], p<0.00005), from 8.4 to a posttest average of 18.5/21 possible points. The FAST knowledge scores increased by a mean of 7.5 points (95% CI [6.3-8.7] p<0.00005), from 8.1 to a posttest average of 15.6/21. We analyzed clinical skills data on 32 students. The mean score was 1.7 pretest and 4.7 posttest of 12 possible points. Mean improvement was 3.0 points (p<0.00005) overall, 3.3 (p=0.0001) for FAST, and 2.6 (p=0.003) for the pulmonary US exam.ConclusionThis study suggests that a symposium on US can improve clinical knowledge, but is limited in achieving image acquisition for pulmonary and FAST US assessments. US training external to official medical school curriculum may augment students' education.
| S-EPMC4307699 | biostudies-literature
Project description:Cancer treatment often lacks individual dose adaptation, contributing to insufficient efficacy and severe side effects. Thus, personalized approaches are highly desired. Although various analytical techniques are established to determine drug levels in preclinical models, they are limited in the automated real-time acquisition of pharmacokinetic profiles. Therefore, an online UHPLC-MS/MS system for quantitation of drug concentrations within 3D tumor oral mucosa models was generated. The integration of sampling ports into the 3D tumor models and their culture inside the autosampler allowed for real-time pharmacokinetic profiling without additional sample preparation. Docetaxel quantitation was validated according to EMA guidelines. The tumor models recapitulated the morphology of head-and-neck cancer and the dose-dependent tumor reduction following docetaxel treatment. The administration of four different docetaxel concentrations resulted in comparable courses of concentration versus time curves for 96 h. In conclusion, this proof-of-concept study demonstrated the feasibility of real-time monitoring of drug levels in 3D tumor models without any sample preparation. The inclusion of patient-derived tumor cells into our models may further optimize the pharmacotherapy of cancer patients by efficiently delivering personalized data of the target tissue.
| S-EPMC7284432 | biostudies-literature
Project description:BackgroundThe field of medical education remains poorly delineated such that there is no broad consensus of articles or journals that comprise 'the field'. This lack of consensus indicates a missed opportunity for researchers to generate insights about the field that could facilitate conducting bibliometric studies and other research designs (e.g., systematic reviews) and also enable individuals to identify themselves as 'medical education researchers'. Other fields have utilised bibliometric field delineation, which is the assigning of articles or journals to a certain field in an effort to define that field.ProcessIn this Research Approach, three bibliometric field delineation approaches-information retrieval, core journals, and journal co-citation-are introduced. For each approach, the authors describe attempts to apply it in medical education and identify related strengths and weaknesses. Based on co-citation, the authors propose the Medical Education Journal List 24 (MEJ-24), as a starting point for delineating medical education and invite the community to collaborate on improving and potentially expanding this list.PearlsAs a research approach, field delineation is complicated, and there is no clear best way to delineate the field of medical education. However, recent advances in information science provide potentially fruitful approaches to deal with the field's complexity. When considering these approaches, researchers should consider collaborating with bibliometricians. Bibliometric approaches rely on available metadata for articles and journals, which necessitates that researchers examine the metadata prior to analysis to understand its strengths and weaknesses, and to assess how this might affect data interpretation. While using bibliometric approaches for field delineation is valuable, it is important to remember that these techniques are only as good as the research team's interpretation of the data, which suggests that an expanded approach is needed to better delineate medical education, an approach that includes active discussion within the medical education community.
| S-EPMC9298433 | biostudies-literature
Project description:How can physicians incorporate the electronic health record (EHR) into clinical practice in a relationship-enhancing fashion ("EHR ergonomics")?Three convenience samples of 40 second-year medical students with varying levels of EHR ergonomic training were compared in the 2012 spring semester. All participants first received basic EHR training and completed a presurvey. Two study groups were then instructed to use the EHR during the standardized patient (SP) encounter in each of four regularly scheduled Doctoring (clinical skills) course sessions. One group received additional ergonomic training in each session. Ergonomic assessment data were collected from students, faculty, and SPs in each session. A postsurvey was administered to all students, and data were compared across all three groups to assess the impact of EHR use and ergonomic training.There was a significant positive effect of EHR ergonomics skills training on students' relationship-centered EHR use (P<.005). Students who received training reported that they were able to use the EHR to engage with patients more effectively, better articulate the benefits of using the EHR, better address patient concerns, more appropriately position the EHR device, and more effectively integrate the EHR into patient encounters. Additionally, students' self-assessments were strongly corroborated by SP and faculty assessments. A minimum of three ergonomic training sessions were needed to see an overall improvement in EHR use.In addition to replication of these results, further effectiveness studies of this educational intervention need to be carried out in GME, practice, and other environments.
| S-EPMC4146679 | biostudies-literature